Big Howe is this why doing these tests on a track instead of the street is more preferred? the ability to obtain more steading readings with more consistent throttle pressures? like you guys were saying....to many variables on the street

Steve

 

I really don't want to jump start the "real world" testing battle again, but, to provide a brief summary.... While running cars on the dyno and track is not perfect, it does minimize many variables, helping one to collect good data. That's not to say one small thing on the dyno could render all runs useless, but it's easier to manage the variables. Testing conditions can be more closely controlled and the human element removed(for the most part, even on the track).
Another important item is a post(#345) by Dr. O earlier in this thread. By testing units to the extreme limits, the differences are usually more glaring. Where you may only see a 2-4F difference on the street, it may be 3 times that at the track. It can be reasonably argued that the differences seen on the track will follow similar trends on the street. You have to watch you margin of error for the test and make sure you data is outside this range as well. There could be special circumstances, for example, how something recovers from heatsoak at standstill, but this can be controlled with a strict set of tests.
All this is not to say what is trying to be accomplished isn't without merit, it's just some items are easier to test and measure than others. The IC in the Mini falls into the harder catagory with the amount of variables and readings involved.

 

If I'm not mistaken, the T-MAP is in the front of the intake manifold.

 

Thats the one I was thinking as well. It will be interesting to log this temp as well as the inlet outlet of the IC.

I'm curious to know how much of an increase in temp there will be from the outlet of the IC to the intake manifold. I'm guessing that it will increase slightly( Well at least I hope that its slightly ) due to the supercharger being underneath it and the radiator hoses running in front of it.

I'm excited to do this It's going to be a lot of work, all in all it will be a fun challenge

Thanks for your willingness to help me out Big H. I really appreciate it

 

I've been busy....

What I do:

1) Set up. Prior to any run, I make sure that IATs (at the T-MAP sensor) are in a small window. I "set" the car by doing a controlled run on a freeway under cruise control. This makes sure the baseline to start my run is constant.

2) Data. You need temps, pressures and RPM at a minimum. (speed is just a function of RPM, so it's really redundant). If you want to add other variables, looking at timing is good (it's easy to see timing pull). Fuel mass is another good one, as you can see if the engine thinks it's getting more air in.

3) Sensors. This is where the testing $ goes. Good ones aren't cheap. FWIW, I was never happy with the sensor set I was running and am making my own. These are products that come out after the shiftlight is releasted to production (250 boards are on thier way to my house this week). FWIW, the time constant on the factory temp sensor is about 10 seconds! That in effect means that the signal you get from the diagnostics port is somewhat smoothed and trunkates peak values.

4) Analysis. This takes a bunch of time. First pass on looking at data is too see what runs are outliers. These are anomolous readings due to lots of stuff: Bad baseline conditioning of the car, too long waiting for a run, sensor moving measurement location etc. About half the runs (maybe a bit less) get thrown out due to various screw ups. It's a royal pain in the butt! One thing that sucked for me was that I was running different loggers with different data rates and time stamps and had to make graphs by hand in Excel. Lots of time down the toilet here.

5) Conclusions. If I'm carefull, I can get a set of readings (3-5 per state) that I keep with a standard deviation of less than 2%. Also, with a 2% SD, that means that anything more than 3 significant figures for any number is pretty much useless (even if impressive). So for HP readings on a car with 250 whp, anything less than about 5 HP delta is pretty much burried in the noise.... You can maybe finess to a bit less than this, but not much. Any delta less than 1% is just meaninless for sure.

So, what have I learned? Good quantitative testing is a royal pain. It's just plain hard to do well. It takes time, lots of it, and good tools. You have to know how the sensors work (like the time constant of the T-MAP sensor mentioned above. FWIW, the pressure part has a 20 millisecond response, I think... That means that the speed/density function doesn't get the best of signals during transients....) to do good analysis. Just sucks, but that's the way it is!

Good luck and happy measuring!

Matt

 

I thank you for your help and knowledge to Matt

I have read your report on the intercooler testing that you did. Some pretty amazing things in there, waaay above my knowledge level for sure.

Basically what your saying here is before each run you get the car to temp and condition it the same before each test correct? I just want to make sure that I am comprehending your information correctly.

So, the testing equipment I have is basic. I can measure temps pre and post IC. I also have a Scanguage II ( Thanks Way Motor works for finding one for me and for rushing it out!)

I can set the Scanguage to read MAP pressure, Intake temp, RPM and Timing. It sounds like there is some lag time so these readings are most likely going to be skewed a bit. Better than nothing I'm guessing though right?

The software that we are using has the ability to write directly to an Excel Spreadsheet. From messing around with it, it seemed to work very well and exported to a graph fairly easy. Basically what the graphs will show is a time stamp on the bottom and temps on the side. We will have to figure out how to work the scanguage info in and match it with the time stamps.

The more I think about doing this the more I realize what a pain in the kiester certain things are going to be. My plan is to do the best analysis as possible and put it all together.

My tools are incomplete but basic baselines can be had
My knowledge of how the sensors work is "lacking" shall we say
My time is minimal but I'm making it happen

I think my drive for knowledge, fun and everything Mini has led me to this adventure. I do think it will be fun, my friend Ryan (Roggenmini) and I have actually been really enjoying learning and experimenting with this stuff.

In short, if the information we provide proves to be absolutely meaningless by the NAM community all won't be lost for us (This is not our hope or our intention. My drive to do this is to provide at least some helpful/meaningful information to those looking to make a decision about IC's). Doing this will still be a fun challenge that at the very least we will enjoy. I know for myself, I have learned an immense amount of useful information about how everything works and fits together on our cars.

One funny thing that might amuse you is that I am struggling a bit with some of the acronyms and terminology people are throwing out

You mentioned "3-5 per state" I'm just guessing that means runs?

"Fuel Mass" I'm not sure what this means either. Fuel flow? How can this be measured?

"I keep with a standard deviation of less than 2%. Also, with a 2% SD, that means that anything more than 3 significant figures for any number is pretty much useless (even if impressive). anything less than about 5 HP delta is pretty much burried in the noise.... You can maybe finess to a bit less than this, but not much. Any delta less than 1% is just meaninless for sure."
I'm not sure what any of this means at all

 

All of this & more. Weather is a huge factor. If you can't repeat, the data is worthless.

 

Well, the weather should be the same but the temps will not. I basically have one maybe two days to complete this in.

So, the ambient temperature will vary through the day

 

When you could be looking at differences of 5-10 degrees(or less) and the ambient changes by 10 degrees, well................

 

Hey maybe you can do something like this: Just modifiy it so you can drive around in it.

 

Holy crap! What the heck is that for?

 

someone really likes their fiero! Ive heard of bubble boys, never bubble cars.

 

It's so you can keep the specs the same for all the tests. Just turn on your A/C and roll the windows down so the inside stays the same temp........... Your going to have to do something for the exhaust tho

 

Laugh all you want -- but 20-30 years from now that Fiero will be worth $5,000

 

For yucks I finally configured the Auterra tool I have to record some IAT's as I ran around town today. Where in the airflow is this measured?

 

Ok,

I think we've narrowed the testbed. It has a combination of freeway, canyon and mild urban driving.

I will mark points on this route when we decide at what points were going to do certain tests.
I couldn't figure out how to link to the Google Earth map I'm working with but you get the idea. I'll try to make this prettier but I thought I'd throw up a rough of where were thinking of running around.

Were still debating this route but so far its the favorite.



http://maps.google.com/maps?ll=40.45...z=13&t=h&hl=en

 

funniest thing Ive read in a while.

 

T-MAP sensor, black box on the front of the intake manifold near the inlet. As Dr. O has pointed out, it does some averaging so you don't see the spikes.
What did you see? What Auterra tool do you have?

 

Don't forget to deduct the cost of the electricity over the 5 years from the $5000.

 

I was going to edit my post for that, but what the hay. I knew someone would bring it up!

 

Have the Dyno-Scan software on my Palm based phone. I've only used to check and clear codes in the past but figured I see what else I can do with it.

Just noodling around and recorded RPM, Timing and IAT.

What I saw is I have lot to learn about how to use it effectively.

Some items I noticed...

Ambient was 92 according to OBC.

Drove to get some crac..err coffee and IAT never went over 120. Normal around town driving. Left crack hou...srr Starbucks and IAT at startup was 148. Normal shopping center traverse and then a quiet cruise down the street. IAT about 115-118 range. Took off in 2nd to 8k IAT got to 151 and was back to 115-118 range within 2 blocks. Pretty much stayed there for majority of sedate cruise home.

I'll play with it more and figure a way to secure it for track session.

 

Sounds very similar to what I'm seeing. And relating to this thread, I'm not sure any intercooler will stop that 151 spike, but the meth/h2o does. You're addicted to the crack too ehhh?

 

Worst gift I was ever given is this damn $100 crack giftcard. Those large doubleshot things where there are 5 and often 6 shots of espresso are worse then heroin.

I really should hook up the water spray, with this hood and a GP back glass I have no use for the washer bottles. I guess I need to start a thread on how to hook them up and when to use along with how to trigger.

 

Hmm. I hope you didn't break the speed limit

 

There's a highway right there JIMINNI......didn't you know.....geeeesh.

Jeremy